Specificity of Trypsin and Chymotrypsin: Loop-Motion-Controlled Dynamic Correlation as a Determinant  Wenzhe Ma, Chao Tang, Luhua Lai  Biophysical Journal  Volume 89, Issue 2, Pages 1183-1193 (August 2005) DOI: 10.1529/biophysj.104.057158 Copyright © 2005 The Biophysical Society Terms and Conditions

Figure 1 Superposition of trypsin and chymotrypsin. (A) The two enzymes have very similar tertiary structure. Trypsin is shown in green ribbon and chymotrypsin in blue. Active site residues of trypsin are shown in ball and stick. Loops of trypsin are shown in magenta; loops of chymotrypsin are shown in pale green. S1 binding pocket is shown in red. This figure is drawn using MOLMOL (67). (B) Sequence alignment of trypsin and chymotrypsin around the L1-L2 loop regions. Black shade indicates loops; gray shade indicates substrate-binding pocket. Lowercase letters represent residues mutated in the experiments. Biophysical Journal 2005 89, 1183-1193DOI: (10.1529/biophysj.104.057158) Copyright © 2005 The Biophysical Society Terms and Conditions

Figure 2 Correlation map of chymotrypsin. Values of correlation between two residues range from −1 to 1. Blue means negative correlation and red means positive correlation, as shown in the color bar on the right. Both x axis and y axis of this map are chymotrypsin residue indices. The two rectangles indicate the relative position of two β-barrels in the protein. Biophysical Journal 2005 89, 1183-1193DOI: (10.1529/biophysj.104.057158) Copyright © 2005 The Biophysical Society Terms and Conditions

Figure 3 Clustering analysis of chymotrypsin. (A) The mean-square fluctuation of each mode. Note the value does not change much after mode 40, so we have used the first 40 modes in the calculation of correlations. (B) The tree of correlations of chymotrypsin. Residues form clusters and we draw a line to define these clusters for the plot in (Fig. 3 C). (C) Different clusters are painted with different colors on the chymotrypsin structure. The colors are chosen arbitrarily. Biophysical Journal 2005 89, 1183-1193DOI: (10.1529/biophysj.104.057158) Copyright © 2005 The Biophysical Society Terms and Conditions

Figure 4 Local correlation trees of chymotrypsin, trypsin, and the hybrid protein. Total length of horizontal lines between two residues is related to the correlation coefficient. The shorter the length, the stronger the two residues are correlated. (A) The local correlation tree of chymotrypsin around the loop regions. Residues on the two loops (●) cluster together with some of the residues in the S1 pocket (◀). (B) The local tree of trypsin. (C) The local tree of the hybrid protein. In all these figures, many residues in the S1 binding pocket cluster with L1-L2 loops. Fig. 4, A and C, are similar in that the correlations between residues 217–219 and L1-L2 loops are stronger in chymotrypsin and the hybrid protein than in trypsin. Residues shown in lowercase letters are those mutated in experiment (1). Figures are drawn by using TreeExplorer (http://evolgen.biol.metro-u.ac.jp/TE/TE_man.html). Biophysical Journal 2005 89, 1183-1193DOI: (10.1529/biophysj.104.057158) Copyright © 2005 The Biophysical Society Terms and Conditions

Figure 5 Comparison of pairwise correlations among residues important for activity. This figure shows S-value of some important residue pairs; x axis entries represent different residue pairs; corresponding y axis entry is the S-value. Most correlations of the hybrid protein are trypsin-like but some correlations between key residues become chymotrypsin-like. Biophysical Journal 2005 89, 1183-1193DOI: (10.1529/biophysj.104.057158) Copyright © 2005 The Biophysical Society Terms and Conditions

Figure 6 Effect of selected modes on protein motion. (A) Contribution of the top modes to the loop region correlation; x axis is mode number, up to 40. Larger numbered modes are not shown because they show little effect on the loop correlation; y axis is the normalized ratio of the contribution. (B) Fluctuations of residues calculated with the most important modes to the loop motion. Modes 3 and 9 were used for trypsin. Modes 3, 4, 5, 6, and 11 were used for chymotrypsin. Modes 3, 4, 5, 9, and 10 were used for the hybrid protein. (C) Mode 3 of the three proteins. (D) Mode 11 in chymotrypsin, mode 10 in the hybrid protein, and mode 9 in trypsin. Biophysical Journal 2005 89, 1183-1193DOI: (10.1529/biophysj.104.057158) Copyright © 2005 The Biophysical Society Terms and Conditions

Figure 7 Correlations near the loop region. Correlations between two residues with an absolute value >0.7 are shown in lines. Correlations between 190–193 and Loop D are shown in red. (A) Chymotrypsin; (B) trypsin; (C) the hybrid protein. In chymotrypsin and the hybrid protein, correlations shown in black are stronger than those in trypsin. Biophysical Journal 2005 89, 1183-1193DOI: (10.1529/biophysj.104.057158) Copyright © 2005 The Biophysical Society Terms and Conditions